DEVELOPMENTAL

BIOLOGY

143,122-129 (1991)

Transgene Expression in the QM Myogenic Cell Line PARKER B. ANTIN, GERALD C. KARP, AND CHARLESP. ORDAHL’ Department

of Anatomy,

University

of California,

San Francisco,

San Francisco,

California

94143

Accepted September 25, 1990 We have isolated an avian muscle cell line (QM) which has the essential features of established mammalian muscle cell lines. The experiments reported here were undertaken to determine the suitability of QM cells for the introduction and analysis of cloned transgenes. The promoter of the cardiac troponin T (cTNT) gene has been previously shown to contain sequence elements which govern muscle-specific expression of the chloramphenicol acetyltransferase (CAT) gene in transiently transfected primary cell cultures. We show here that QM cells stably harboring cTNT promoter-CAT fusion genes up-regulate CAT expression in concert with myogenic differentiation, and that as few as 110 upstream nucleotides are sufficient for such differentiation-dependent regulation. In addition, both transient and stable transfection experiments demonstrate that differentiated QM cells possess tram-acting factors necessary for the expression of the skeletal a-actin promoter, despite the absence of mRNA or protein product from the endogenous sarcomeric aetin genes in these cells. Finally, to follow the developmental potential of QM cells in Go, we created a clone, QMBADH, which constitutively expresses the histochemical marker transgene encoding Drosophila alcohol dehydrogenase. When surgically inserted into the limb buds of developing chick embryos, QMZADH cells are incorporated into endogenous developing muscles, indicating that QM cells are capable of recognizing and responding to host cues governing muscle morphogenesis. Thus, QM cells are versatile as recipients of transgenes for the in vitro and in viva analysis of molecular 0 1991 Academic Press, Inc. events in muscle development.

previously emerged. Such a cell line could be genetically modified via the stable introduction of cloned genes into Experimentation using avian embryos has yielded a their chromosomal DNA, an approach not feasible with substantial fraction of our current knowledge regarding primary cells. Although stable transfection has been exembryonic muscle development. The first explant cul- tensively used to study the differentiation-dependent tures of striated muscle were derived from avian em- regulation of muscle genes in mammalian muscle cell bryos over ‘70years ago (Lewis, 1915; Lewis and Lewis, lines, not all avian muscle genes are correctly regulated 1917) and today, cultured cells isolated from avian em- in a mammalian muscle cell background (Seiler-Tuyns bryos continue to be extensively used to analyze the cel- et al., 1984; Nude1 et al, 1985; Sharp et aZ.,1987). Stable lular and molecular events of myogenesis in vitro (Kon- transfection could also be used to genetically modify igsberg, 1979). The origin of embryonic lineages giving muscle cells for reintroduction into developing embryos rise to the myogenic and fibrogenic components of mus- to study aspects of myogenesis which cannot be reprocle tissue was determined using chimeras between chick duced in vitro. Such an approach is relatively straightand quail embryos (Chevallier et aL, 1977; Jacob et al., forward using avian embryos, which are accessible to 1979). Wolpert and collaborators used surgical tech- surgical intervention throughout most of embryonic niques on avian embryos to clarify the morphogenetic and fetal development. interrelationships between muscle and nonmuscle tisIn the accompanying paper (Antin and Ordahl, 1991), sues (Shellswell and Wolpert, 1977; Lewis et al., 1981). we describe a quail muscle cell line (QM) whose characMore recently, the elaboration of specialized lineages teristics closely resemble those of mammalian muscle within developing muscle has been elucidated using cell lines. In the present paper, we report experiments in myogenic cells isolated from developing avian embryos which QM cells have been genetically modified via the (Bonner and Hauschka, 1974;Miller and Stockdale, 1986; stable introduction of cloned genes. In one set of experiSchafer et al, 1987). ments we analyze, for the first time, the differentiationDespite the extensive research on avian muscle em- dependent regulation of the cardiac troponin T gene bryogenesis, a stable, immortal avian myogenic cell line promoter. We also analyze the regulation of the skeletal suitable for molecular biological analyses has not a-actin promoter, which is of particular interest because QM myotubes do not express cu-actin mRNA or protein. In a second set of experiments we show that QM ’ To whom correspondence should be addressed. cells stably harboring a marker gene encoding a histoINTRODUCTION

0012-1606/91 $3.00 Copyright All rights

Q 1991 by Academic Press, Inc. of reproduction in any form reserved.

122

ANTIN,

KARP,

AND ORDAHL

chemically stainable product can participate in normal muscle development when implanted into developing chick embryo limb buds. Thus, QM cells may prove useful for the molecular analysis of muscle differentiation both in vitro and in viva. MATERIALS

AND

METHODS

Cell Culture and Stable Transfection

of QM Cells

Isolation and characterization of the avian QM cell line has been described (Antin and Ordahl, 1991). Two QM subclones, designated QM2 and QM7, were used for all experiments reported here. QM cells were maintained as replicating precursor cells on uncoated plastic tissue culture dishes (Corning) in growth medium (Ml99 medium plus Earles balanced salt solution [GIBCO, Grand Island, NY] supplemented with 10% fetal calf serum, 10% tryptose phosphate, glutamine, 100 U/ml of penicillin, and 100 pg/ml streptomycin). To induce differentiation, QM cells were grown to confluence and were then shifted to differentiation medium (Ml99 medium plus Earles balanced salt solution, supplemented with 0.5% fetal calf serum, penicillin, and streptomycin) for various lengths of time. Cultures of Day 9 quail embryo pectoralis were prepared as described (Konigsberg, 1979). QM cells were stably transfected by the calcium phosphate method (Gorman et al., 1982; Gorman, 1985) using 40 pg of target DNA (RSVADH, cTNT-550-CAT, cTNTllO-CAT, sACT-2000-CAT, or P-a&in-CAT; see Results) and 0.4 pg of SVBNeo, a plasmid containing the neomycin resistance gene (Neo) under control of the SV-40 promoter (Southern and Berg, 1982). Forty-eight hours after transfection, cells were subcultured at 5 X lo5 cells per loo-mm dish into growth medium containing 400 pg/ml Geneticin (GIBCO). After 7 days, Geneticin was removed from the culture medium and surviving cells were maintained in growth medium mixed 1:l with medium which had been previously conditioned by passage over QM cells in log-phase growth for 24 hr. A QM2 clone containing RSVADH was identified and subcloned twice. The resulting clone, which expresses high levels of Drosophila alcohol dehydrogenase (dADH), was named QMBADH. For stable transfection analysis of promoter-CAT constructs, more than 200 individual neomycin-resistant QM clones were isolated and pooled for each transfected plasmid. Chloramphenicol acetyltransferase assays were performed as described elsewhere (Mar et al., 1988). Transient transfection analysis was performed as previously described (Mar et al., 1988). Injection

sf QM Cells into Embryos

QMBADH trypsinization

cells in growth medium were collected by and centrifugation and resuspended at

Tramgenie

123

QM Cells

approximately 1 x lo4 cells per microliter in ice-cold Howards Ringer containing 10 mg/ml Dextran Blue (Sigma Chemical Co., St. Louis, MO) as a marker dye. Injection pipettes were prepared from capillary tubes (No. IBlOO-4, World Precision Instruments Inc., New Haven, CT) using a Narshige PW-6 needle puller and backfilled with cell suspension using a lo-p1 Hamilton syringe. Stage 22-28 embryo limb buds (Hamburger and Hamilton, 1951) were then injected with l-2 ~1 of the cell suspension using a Picospritzer II injector (General Valve Corp., Fairfield, NJ). Histochemistry Injected limbs of the appropriate age were placed in OCT (Miles Inc., Elkhart, IN) and frozen in melting freon for 15 see before transfer to liquid nitrogen. Frozen serial sections were routinely obtained from several areas of each limb and were then placed on adjacent slides for enzymatic or immunofluorescence analysis. Enzymatic staining for Drosophila ADH was performed as described (Ursprung et al., 1970), with the modifications described below. Slides containing frozen sections were placed in phosphate-buffered saline (PBS) for 5 min, then fixed for 10 min in 2.5% formaldehyde at room temperature. Sections were then washed three times in PBS and incubated in staining solution composed of 175 mMTris-HCl (pH 8.2), 0.3 MB-butanol, 0.44 mg/ml nitro blue tetrazolium (Sigma), 0.017 mg/ml phenazine methosulfate (Sigma), and 0.44 mg/ml NAD+ (Sigma) for 1-3 hr in the dark at room temperature. After light counterstaining with eosin, sections were dehydrated and mounted in Permount. Double-label immunofluorescence was performed essentially as described elsewhere (Antin ef al., 1988; Mar et al., 1988). Antisera prepared against dADH was the gift of W. Sofer (Rutgers University). Antibody against light meromyosin was the gift of H. Holtzer (University of Pennsylvania). RESULTS

AND

DISCUSSION

QM Cells Regulate u Transfected Muscle-SpeciJc Promoter in a Differentiation-Dependent Mann,er Transient transfection experiments using primary chicken muscle cultures have shown that chimeric gene constructions containing promoter/upstream segments of the cardiac troponin T (cTNT) gene coupled to the bacterial chloramphenicol acetyltransferase (CAT) gene are regulated in a cell-specific manner (Mar et al., 1988; Mar and Ordahl, 1988, 1990). However, because differentiation of the primary muscle cells employed in those experiments cannot be controlled via medium conditions, it was not feasible to determine if transcription

124

DEVELOPMENTALBIOLOGY

-500 4

-400

-300

a

-200

:100 +-b+

+l

b

FIG. 1. Cardiac troponin T promoter constructions used for stable transfection into QM2 cells. The upper part of the figure shows regulatory domains of a chicken cardiac troponin T 5’ flanking region previously shown to be important for its expression in skeletal muscle cells (Mar et ab, 1988; Mar and Ordahl, 1988, 1990). Arrows a and b show the promoter segments used to direct CAT expression in cTNT550-CAT and cTNT-110-CAT, respectively. The dashed line indicates the upstream limit of cTNT-llO-CAT. The approximate positions of the muscle-specific enhancer (diagonal stripes), CArG, MEF 1, MCAT, and TATA motifs are indicated. The filled box indicates the cTNT promoter segment which contains muscle-specific transcriptional determinants. Note that neither the enhancer nor the CArG motif is contained within cTNT-llO-CAT. Details regarding these constructions can be found in Mar et al. (1988).

of the chimeric cTNT-CAT trams genes was activated in a differentiation-dependent fashion. Drugs which block myogenic differentiation have been used to analyze the differentiation dependence of transgene activation in primary muscle cultures (Billeter et ab, 1988; Arnold et ab, 1988). However, since QM cell differentiation can be controlled via medium conditions, these cells afforded us the opportunity of determining the differentiation dependence of cTNT promoter activation in the absence of drugs. Transgenes containing the CAT reporter gene under transcriptional control of the cTNT or P-actin gene promoters (see Fig. 1 and Mar et al., 1988) were transfected into QM cells along with the selectable marker plasmid SVBNeo. More than 200 neomycin-resistant clones for each construction were pooled and expanded for analysis of CAT expression during myogenic differentiation. Cultures of pooled clones were grown to confluence in growth medium, switched to differentiation medium, and then sacrificed at various time intervals for analysis of CAT activity. Under these conditions, fusion was evident 24 hr following medium shift, and by 96 hr cultures contained large numbers of multinucleated myotubes (Antin and Ordahl, 1991). Cultures containing pactin-CAT showed little or no change in the relative amount of CAT expression after medium shift (Fig. 2, open squares). This result is in agreement with other reports showing that down-regulation of P-actin gene transcription during myogenesis required sequences in the 3’ end of the gene (DePonti-Zilli et al, 1988). In contrast, CAT expression under the control of cTNT-550 (which contains 550 nucleotides upstream of the transcription initiation site; see Fig. 1) increases over 30-fold following medium shift (Fig. 2, open circles). This in-

VOLUME 143.1991

crease closely follows the increase in myocyte fusion observed microscopically (Antin and Ordahl, 1991). We conclude, therefore, that transcription of the transfected cTNT promoter is up-regulated in a differentiation-dependent fashion in QM cells. Cell-specific expression of the cTNT promoter is dependent upon two tandemly repeated heptanucleotide M-CAT sequences (5’-CATTCCT-3’) contained within the first 100 nucleotides upstream of the cTNT transcription initiation site (see Fig. 1, and Mar and Ordahl, 1988,199O). To determine if a “minimal” cTNT promoter is sufficient to activate CAT expression in a differentiation-dependent manner, cTNT-llO-CAT was cotransfected into QM cells along with the SVBNeo selectable marker as described above. Figure 2 (closed circles) shows that CAT expression under control of cTNT-110 is up-regulated in parallel with that observed for cTNT550-CAT. We conclude that 110 upstream nucleotides are sufficient to up-regulate transcription of the cTNT promoter in a differentiation-dependent fashion. The region between -550 and -110, which has been deleted in cTNT-110, contains two sequence elements of interest for this experiment. First, at position -120 is the sequence 5’-CCAAATAGC-3’, which closely resembles the CArG motif previously shown to be essential for skeletal and cardiac cu-actin gene expression in skeletal muscle cells. Recent experiments show that the CArGlike motif is not required for muscle-specific expression of the cTNT promoter, although nuclear protein binding to this region can be detected by footprinting (Mar and Ordahl, 1990). Since this sequence is absent from cTNT-

0

60

24

96

Hours FIG. 2. Time course of CAT expression in differentiating QM cells. Cultures of pooled QM2 clones containing stably integrated fl-actinCAT (open squares), cTNT-550~CAT (open circles), cTNT-llO-CAT (closed circles), or pooled QM7 clones of sACT-2000-CAT (open triangles) were switched from growth medium to differentiation medium at time zero and then sacrificed at the indicated times. Extracts from equal numbers of cells were analyzed for CAT activity. Under these culture conditions fusion occurred between 24 and 48 hr after medium shift. SEM were smaller than the symbols used for data points.

110, we conclude that it is also not required for differentiation-dependent activation of the cTNT promoter. A second sequence element which is apparently not required for differentiation-dependent activation is a muscle-specific enhancer located between nucleotides -550 and -268 (see Fig. 1 and Mar et ah, 1988). We note, however, that the relative increase of cTNT-110 is approximately threefold lower than that of cTNT-550, consistent with previous findings indicating that the upstream enhancer increases expression of the minimal promoter approximately threefold (Mar et al., 1988). Thus, the cTNT enhancer may augment the differentiation-dependent up-regulation of the cTNT promoter in a quantitative manner. cTNT-110 contains a MEF 1 motif (Buskin and Hauschka, 1989) and two M-CAT motifs (Mar and Ordahl, 1988,1990), both of which have been implicated in regulation of muscle gene expression. Muscle-specific expression of the cTNT promoter is absolutely dependent upon the M-CAT motifs, while the MEF 1 motif is dispensible. Further work will be necessary to determine if the requirements for differentiation-dependent up-regulation are the same as, or different from, those required for muscle-specific expression. QM Cells Express a Transfected

a-Actin

Promoter

125

Transgenic QM Cells

ANTIN,KARP,ANDORDAHL

A Hind

Hind

III

Ill I

6

Hind (

Hind

Ill //“,”

c,

-2000

-200

Ill

C.TH -100

+l

+27

C

*UC

“A0

sACT-2000-CAT

FIG. 3. Construction of sACT-2000-CAT. (A) Structure of the Hind111 fragment containing the skeletal oc-actin (sACT) gene in clone pGa-actin 1 (Ordahl et al, 1980; Fornwald et al, 1982). Open boxes represent the seven exons of the skeletal actin gene; thin lines represent introns and flanking sequence. The double-headed arrow indicates the subcloned promoter upstream region shown in (B). (B) The skeletal a-actin promoter upstream region between the natural Hind111 site 2000 nucleotides upstream and an artificial Hind site inserted into exon I, 27 nucleotides downstream of the transcription initiation site. The 3’ end truncation was generated by Bal31 digestion from a Dde 1 site located in intron I by methods described elsewhere (Mar and Ordahl, 1988). The exact deletion point was determined by sequencing. (C) The region shown in (B) was subcloned into the Hind111 site of pBRCAT (Walker et al., 1983) to create sACT-2000CAT. Filled box; coding sequence of the CAT gene with arrow showing the direction of transcription. AUG and termination codons are indicated. Hatched box; SV40 splice signal and polyadenylation sequence. The thin line represents pBR322 sequence. Ori, pBR322 origin of replication AmpR, ampicillin resistance gene. Abbreviations: T, TATA; M, M-CAT; C, CarG/CBAR homologies; U, evolutionarily conserved upstream A/T rich regions (adapted from Chow and Schwartz, 1990).

In the accompanying paper we demonstrate that differentiated QM cells do not contain sarcomeric actin protein (Antin and Ordahl, 1991). The molecular basis for this is unknown, but the absence of sarcomeric actin mRNA suggests that it may result from a transcriptional or post-transcriptional defect. We sought, therefore, to determine if QM cells can express and regulate a transfected cu-actin promoter functionally linked to a CAT reporter gene. A segment of the chick skeletal (Yactin gene extending from a natural Hind111 site approximately 2000 nucleotides upstream, to an artificial Hind111 site created 27 nucleotides downstream, of the transcriptional initiation site at +l, was inserted into the Hind111 site of pBR-CAT to create sACT-2000-CAT parable to that of cTNT-550-CAT (hatched bars), al(Fig. 3). This promoter contains the upstream sequences though the latter shows a higher overall level of activity previously shown to be sufficient to direct cell-specific than sACT-2000-CAT (Fig. 4). In contrast, the P-a&in expression in cultures of chick primary muscle cells promoter (open bars) is active in all three cell types, as (Grichnik et al., 1986). expected for this non cell-specific promoter (Mar et al., To assess the cell-specific activity of sACT-2000-CAT 1988). in differentiated QM cells, we compared its activity in To determine whether a transfected cr-actin promoter transiently transfected myogenic and nonmyogenic would be correctly regulated during differentiation of cells. Figure 4 shows that sACT-2000-CAT (black bars) QM cells, we created pooled QM7 clones carrying stably is as active in differentiated cultures of QM7 cells as in transfected sACT-2000-CAT. Figure 2 shows that the differentiated cultures of chick primary breast muscle. CAT expression in these pooled clones increases approxiHowever, sACT-2000-CAT is essentially inactive in culmately threefold within 96 hr after shift to differentiatures of QM5 cells, a nonmyogenic QM derivative (Antin tion medium. This degree of induction is less than that and Ordahl, 1991). This pattern of cell specificity is com- observed with either cTNT-110 or cTNT-550 (Fig. 2 and

126

DEVELOPMENTAL BIOLOGY

Breast Muscle FIG. 4. Activity of /3-actin-CAT (open bars), cTNT-550-CAT (hatched bars), and sACT-ZOOO-CAT (black bars) in differentiated cultures of Day 11 chick embryo pectoralis and QM7, and in the nondifferentiating QM variant, QM5. Pectoralis cultures were transfected on Day 1 following plating and maintained for an additional 4 days in standard differentiation medium (see Materials and Methods). Cultures of QM7 and QM5 in log-phase growth were switched t.o differentiation medium 24 hr following transfection and were fed daily until sacrifice on Day 4. Differentiated pectoralis and QM7 cultures contained large numbers of multinucleated myotubes. QM5 cells failed to fuse in differentiation medium. Data are shown in CAT units, error bars denote SEM.

see above), and also less than predicted from earlier experiments with similar constructions of this promoter from birds and mammals (Bergsma et ab, 1986; Muscat and Kedes, 1987). Nevertheless, that bona fide induction of sACT-2000-CAT occurs is evidenced by comparison to bACT-CAT, which is clearly not induced after medium shift (Fig. 2). We conclude from these experiments that QM muscle cells possess trans-acting factors necessary for transcription of the skeletal cli-actin promoter. This suggests that the absence of appropriate transcriptional factors per se is unlikely to completely explain the absence of sarcomeric actin mRNA and protein in QM muscle cells. On the other hand, the poor induction of a transfected skeletal cu-actin promoter in QM cells after medium shift indicates possible defects in the developmental regulation of this promoter which may indeed be related to the absence of this gene product in QM cells. Thus, different components of actin gene regulation may be uncoupled in QM cells. QM Cell Participation in Muscle Fiber Formation Developing Limb Buds

in

While cultured muscle cells reiterate many essential features of myogenesis, there are a number of important features of muscle development that do not occur in culture. Many of the isogene switches that are observed in vivo, for example, do not occur under standard in vitro conditions. Additionally, many aspects of cell-cell and

VOLUME 143,199l

cell-matrix interactions which are important for muscle development and maturation are difficult to reproduce in culture. Ideally, it would be interesting to combine in vitro and in vivo approaches to the analysis of muscle development by introducing genetically modified cell lines into developing embryos. To determine if QM cells are suitable for such a purpose we introduced genetically marked QM cells into the limb buds of developing chick embryos. It was necessary to genetically mark QM cells for this experiment because quail cell lines propagated in vitro lose the nucleolar markers which have proven so valuable for chickquail chimera experiments (N. Le Douarin, personal communication). We therefore created a QM cell line stably expressing dADH under control of the rous sarcoma virus long terminal repeat (RSV-ADH; Ordahl et al., 1986). This cell line, designated QMBADH, constitutively expresses dADH, which can be distinguished histochemically from vertebrate ADH isozymes by virtue of its ability to use butanol as a substrate. In a typical embryo injection experiment, approximately 1 X lo4 QMBADH cells from log-phase cultures were injected into the wing and leg buds of 14 stage 22-28 chicken embryos. These stages of chick development span the period of major muscle morphogenesis in the limb buds. Embryos were sacrificed 4-7 days later and limbs were processed for histochemistry. Six of nine surviving embryos had dADH-positive cells in their developing limbs. Figure 5A shows a section through the wing of an embryo 6 days following injection of QMBADH cells. An opaque enzymatic reaction product reveals broad distribution of QMZADH cells within a muscle mass in the center of the wing (arrow). Serial sectioning showed that the dADH reaction product was extensively distributed within this muscle mass, extending through more than one hundred lo-pm sections. Figure 5B shows a higher magnification view of dADH staining within the muscle mass of a second injected limb. In this section the reaction product appears localized within three distinct myofibers of the developing muscle. Thus, QM cells are capable of extensively populating muscle tissue within the developing chick limb. To verify that the dADH product was localized within authentic muscle, wing sections were double stained with antibodies against dADH and vertebrate myosin heavy chain. Figure 5C shows anti-myosin heavy chain staining of several muscle fibers within a muscle mass of a typical experimental limb section. Figure 5D shows anti-dADH binding within the central muscle fiber (see arrow) of this group. Thus, QM cells can participate in the development of authentic muscle. Rarely, dADHpositive cells were found scattered singly within the limb. These isolated cells were usually myosin heavy

ANTIN, KARP, AND ORDAHL

Trunsgenic

QM Cells

FIG. 5. (A, B) Bright-field micrographs of sections through Day 10 wings, 6 days following injection of QMSADH cells. (A) Low magnification view of a wing showing an opaque ADH reaction product within a central muscle mass; bar, 1 mm. (B) Higher magnification view of a section from a second wing showing three ADH-positive myotubes within a developing muscle; bar, 100 Frn. (C, D) Immunofluorescence micrographs of the same microscopic field, visualizing myosin heavy chain in (C) and ADH in (D). Arrows point to a myofiber staining with both antibodies; bar, 100 pm. (E, F) Immunofluorescence micrographs of the same microscopic field, showing a single myofiber stained with myosin heavy chain in (E) and ADH in (F); bar, 50 pm. (G) Higher magnification of inset from (E) clearly showing a striated myosin staining pattern; bar, 20 pm.

chain negative, indicating that they had not undergone myogenic differentiation. In addition, round aggregates of dADH-myosin heavy chain-positive cells were infrequently observed at ectopic sites. To estimate the extent of ectopic versus homotopic QM differentiation, 14 additional injected limbs were examined. In these specimens, essentially all of the dADH-myosin heavy chainpositive fibers observed were within or very near endog-

enous muscle masses. This indicates that transplanted QM cells tend to differentiate in close proximity of, and in the appropriate orientation relative to, endogenous host myofibers. To determine if such dADH-positive muscle fibers constitute heterokaryons between QM2 and endogenous chicken muscle cells, we took advantage of the fact that cultured QM muscle cells cannot independently assem-

128

DEVELOPMENTALBIOLOGY

ble striated myofibrils (Antin and Ordahl, 1991). Therefore, localization of dADH within cross-striated myofibers should positively identify true heterokaryons. Figures 5E and 5F show a group of myosin heavy chainpositive fibers (Fig. 5E) which also bind anti-dADH (Fig. 5F). At higher magnification the anti-myosin heavy chain staining of this fiber (Fig. 5G) shows a cross-striated banding pattern characteristic of fully assembled myofibrils. Although visualization of striations depended upon precise orientation of muscle fibers, a banded myosin heavy chain pattern was observed in several ADH-positive fibers. We conclude that QM cells injected into the limb bud can form heterokaryons at high frequency with endogenous chicken muscle cells. The above results demonstrate that QM cells injected into developing chick embryos not only survive and differentiate but can participate in the formation of authentic muscle fibers. Moreover. the close colocalization, coorientation, and cofusion of QM cells within endogenous muscle fibers suggest that the QM cells are capable of responding to the endogenous host cues directing muscle morphogenesis. SUMMARY AND CONCLUSIONS

The experiments reported here demonstrate that QM cells are amenable to stable genetic modification via transfection of cloned DNA encoding selectable and nonselectable markers without diminution of their myogenic potential. These properties are commonplace among mammalian muscle cell lines, and QM cells now allow such experiments to be performed in an avian cell background. We have used QM cells to analyze aspects of the developmental regulation of an avian muscle gene which might otherwise be difficult or impossible to study. We show that the sequences required for cell-specific expression of the cTNT promoter are also sufficient for its up-regulation during the developmental transition from replicating myoblast to myotube. In the case of the skeletal a-actin promoter, we show that although QM cells are capable of expressing this promoter in a cell-specific fashion, they do not up-regulate this promoter to the extent expected during differentiation. This apparent defect partially explains the absence of endogenous sarcomeric actin mRNA or protein in differentiated QM cells. Finally, we demonstrate that genetically modified QM cells can participate in apparently normal muscle differentiation and morphogenesis irL vivo, after surgical implantation into developing chick embryos. Thus, avian myogenic cell lines such as QM may prove useful for studying avian muscle development both in vivo and in vitro. We thank our colleagues for generously sharing useful advice and equipment for the pursuit of these experiments. We also acknowledge

VOLUME143. 1991

the expert technical assistance of Nina Kostanian. This work was supported by USDA Grant 87CRCR-12404and NIH Grants HL35561 and

GM32018. REFERENCES ANTIN, P. B., MAR, J. H., and ORDAHL,C. P. (1988). Single cell analysis of transfected gene expression in primary heart cultures containing multiple cell types. BioZ’echniques 6,640-648. ANTIN, P. B., and ORDAHL,C. P. (1991). Isolation and characterization of an avian myogenic cell line. Dev. Biol. 143. ARNOLD, H. H., TANNICH, E., and PATERSON,B. M. (1988). The promoter of the chicken cardiac myosin light chain 2 gene shows cellspecific expression in transfected cultures of chicken muscle. Nucleic Acids Rex 16, 2411-2429. BERGSMA,D. J., GRICHNIK,J. M., GOSSETT,L. M., and SCHWARTZ,R. J. (1986). Delimitation and characterization of cis-actin DNA sequences required for the regulated expression and transcriptional control of the chicken skeletal a-actin gene. Mol. Cell. Biol. 6,24622475. BILLETER, R., QUITSCHKE,W., and PATERSON,B. M. (1988). Approximately 1 kilobase of sequence 5’ to the two myosin light-chain lf/3f gene cap sites is sufficient for differentiation-dependent expression. Mol. Cell. Biol. 8, 1361-1365. BONNER,P. H., and HAUSCHKA,S. D. (1974). Clonal analysis of vertebrate myogenesis. I. Early developmental events in the chick limb. Lkv. Biol. 37, 317-328. BUSKIN,J. N., and HAUSCHKA,S. D. (1989). Identification of a myocyte nuclear factor that binds to the muscle-specific enhancer of the mouse muscle creatine kinase gene. Mol. Cell. BioL 9,2626-2640. CHEVALLIER,A., KIENY, M., and MAUGER,A. (1977).Limb-somite relationship: Origin of the limb musculature. J. EminyoL Ezp. Morphol. 41,245-258. CHOW,K.-L., and SCHWARTZ,R. J. (1990). A combination of closely associated positive and negative cis-acting promoter elements regulates transcription of the skeletal a-actin gene. Mol. Cell. Biol. 10(2), 528-538. DEPONTI-ZILLI, L., SEILER-TUYNS,A., and PATERSON,B. M. (1988). A 40-base-pair sequence in the 3’ end of the @actin gene regulates fl-actin mRNA transcription during myogenesis. Proc. N&L Acad. Sci. USA 85,1389-1393.

FORNWALD,J. A., KUNCIO,G., PENG,I., and ORDAHL,C. P. (1982). The complete nucleotide sequence of the chick a-actin gene and its evolutionary relationship to the actin gene family. Nucleic Acids Res. 10(13), 3861-38’76. GORMAN,C. M. (1985). High efficiency gene transfer into mammalian cells. In “DNA Cloning” (D. M. Glover, Ed.), pp. 143-190. IRL Press, Washington, DC. GORMAN,C. M., MOFFAT,L. F., and HOWARD,B. H. (1982). Recombinant genomes which express chloramphenicol acetyltransferase in mammalian cells. Mol. Cell. Biol. 2, 1044-1051. GRICHNIK, J. M., BERGSMA,D. J., and SCHWARTZ,R. J. (1986). Tissue restricted and stage specific transcription is maintained within 411 nucleotides flanking the 5’ end of the chicken a-skeletal actin gene. Nucleic Acids Res. 14, 1683-1701. HAMBURGER, V., and HAMILTON, H. L. (1951). A series of normal stages in the development of the chick embryo. J. Mmphol. 88,49-92. JACOB,M., CHRIST,B., and JACOB,H. J. (1979). The migration of myogenie cells from the somites into the leg region of avian embryos. Awt. Embryo1 157, 291-309. KONIGSBERG,I. R. (1979). Skeletal myoblasts in culture. I71 “Methods in Enzymology” (W. B. Jakoby and I. H. Pastan, Eds.), Vol. 58,511527. Academic Press, San Diego. LEWIS,J., CHEVALLIER,A., KIENY, M., and WOLPERT,L. (1981). Muscle

ANTIN, KARP, AND ORDAHL nerve branches do not develop in chick wings devoid of muscle. J. Exp Mcrrphol. 64, 211-232. LEWIS, M. R. (1915). Rhythmical contraction of the skeletal muscle tissue observed in tissue cultures. Amer. J. Physiol. 38, 154-161. LEWIS, W. H., and LEWIS, M. R. (1917). Behavior of cross striated muscle in tissue culture. Amer. J. Anat. 22, 169-194. MAR, J. H., ANTIN, P. B., COOPER,T. A., and ORDAHL, C. P. (1988). Analysis of the upstream regions governing expression of the chicken cardiac troponin T gene in embryonic cardiac and skeletal muscle cells. J. Cell Biol. 107.573-585. MAR, J. H., and ORDAHL,C. P. (1988). A conserved CATTCCT motif is required for skeletal muscle-specific activity of the cardiac troponin T gene promoter. Proc. N&l. Acud. Sci. USA 85,6404-6408. MAR, J. H., and ORDAHL, C. P. (1990). M-CAT binding factor, a novel transacting factor governing muscle-specific transcription. Mol. Cell. Biol. 10, 42’7-4283. MILLER, J. B., and STOCKDALE,F. E. (1986). Developmental regulation of the multiple myogenic cell lineages in the avian embryo. J. CeU Biti. 103, 21972208. MUSCAT,G. E. O., and KEDES, L. (1987). Multiple 5’-flanking regions of the human a-skeletal actin gene synergistically modulate musclespecific expression. Mol. Cell. Biol. 7, 4089-4099. NUDEL, U., GREENBERG,D., ORDAHL,C. P., SAXEL,O., NEUMAN,S., and YAFFE, D. (1985). Developmentally regulated expression of a chick muscle-specific gene in stably transfected rat myogenic cells. Proc. Endvyol.

NutL Acud. Sci. USA 82, 3106-3109.

ORDAHL, C. P., TILGHMAN, S. M., OVITT, C., FORNWALD,J., and LARGEN, M. T. (1980). Structure and developmental expression of the chick tu-actin gene. Nucleic Acids Res. 8(21), 4989-5005. ORDAHL, C. P., MAR, J., CLYNE, J., and SIMPSON,P. (1986). Gene

Trclr/Lsyc~?l ic QM Cells

129

transfer into myogenic cells in primary culture. I?r “Molecular Biology of Muscle Development” (C. Emerson, I). Fischman, 8. NadalGinard, and M. A. Q. Siddiqui, Eds.), pp. 547-558. A. R. Liss. New York. SCHAFER,D. A., MILLER, J. B., and STOCKDALE,F. E. (1987). Cell diversification within the myogenic lineage: In vitro generation of two types of myoblasts from a single myogenic progenitor cell. Cell 48, 659-670. SEILER-TUYNS,A., ELDRIDGE,J. D., and PATERSON,B. M. (3984). Expression and regulation of chicken actin gene introduced into mouse myogenic and nonmgogenic cells. Proc. Natl. Acccd. Sci. c:i’A 81, 2980-2984. SHARP, S. B., KOST,T. A., HUGHES, S. H., ORDAHL, C. P., and DAVIDSON,N. (1987). Regulation of intact and hybrid p and (Yactin genes inserted into myogenic cells. In “Molecular Approaches to Developmental Biology” (R. A. Firtel and E. H. Davidson, Eds.), pp. 609-618. A. R. Liss, New York. SHELLSWELL,G. B., and WOLPERT,L. (1977). The pattern of muscle and tendon development in the chick wing. 1~ “Vertebrate Limb and Somite Morphogenesis” (D. A. Ede, J. R. Hinchliffe, and M. Balls, Eds.), pp. 71-86. Cambridge Univ. Press, Cambridge. SOUTHERN,P., and BERG, P. (1982). Transformation of mammalian cells to antibiotic resistance with a bacterial gene under control of the SV40 early region promoter. J. MO/. AppL Genet 1, 327-341. URSPRUNG,G., SOFER,W. H., and BURROUGHS, N. (1970). Tissue distribution of alcohol dehydrogenase in Drosophila melanogastrr. Hi’/helm Roux Arch. 164, 201-208. WALKER, M. D., EDLUND,T., BOULET,A. M., and RUTTER,W. J. (1983). Cell-specific expression controlled by the flanking region of insulin and chymotrypsin genes. Nature Ilondon) 300, 557-561.